Attribution of Surface Radiation Biases in NWP and Climate Models near the US Southern Great Plains

Attribution of Surface Radiation Biases in NWP and Climate Models near the US Southern Great Plains Kwinten Van Weverberg, Cyril Morcrette, Hsi-Yen Ma, J. Petch, S.A. Klein, C. Zhang, S. Xie, Q. Tang, W.I. Gustafson Jr., Y. Qian, L.K. Berg, Y. Liu, M. Huang, M. Ahlgrimm, R. Forbes, E. Bazile, R. Roehrig, J. Cole, W. Merryfield, W.-S. Lee, F. Cheruy, L. Mellul, Y.- C. Wang, K. Johnson, M. Khaiyer 5 th WGNE Workshop on Systematic Errors www.metoffice.gov.uk 19 June 2017, Montreal Crown Copyright 2016, Met Office

Midlatitude Continental Warm Bias 2m-Temperature Bias (K) Colours: Dots: Ma et al. JC 2014 CMIP5 ensemble for 20 years Day 5 forecast using same GCMs

Midlatitude Continental Warm Bias Hypotheses: Soil-vegetation-atmosphere Boundary-layer clouds Convective storms

CAUSES Clouds Above the US and Errors at the Surface Research Foci Radiation errors particularly due to clouds Led by U.K. Met Office Precipitation and surface energy budget errors Led by U.S. Department of energy, LLNL Day 1 Day 2 Day 3 Day 4 Forecast Time Day 5 5-day hindcasts for every day 1 April 31 August 2011 http://portal.nersc.gov/project/capt/causes/

CAUSES: Temperature Bias Clouds Above the US and Errors at the Surface METUM CAM5 TaiESM IFS WRF-CLM WRF-NOAH CNRM LMDZ CanCM4 Day 2-5 mean temperature bias April August 2011 Morcrette et al. 2017 JGR (submitted)

CAUSES: Temperature Bias Clouds Above the US and Errors at the Surface METUM CAM5 TaiESM IFS WRF-CLM WRF-NOAH CNRM LMDZ CanCM4 Is ARM Southern Great Plains site representative? Morcrette et al. 2017 JGR (submitted)

CAUSES: Temperature Bias Clouds Above the US and Errors at the Surface METUM CAM5 TaiESM IFS WRF-CLM WRF-NOAH CNRM LMDZ CanCM4 Correlation of the temperature bias with the SGP-site Morcrette et al. 2017 JGR (submitted)

CAUSES: Radiation Attribution Clouds Above the US and Errors at the Surface

CAUSES: Radiation Attribution Clouds Above the US and Errors at the Surface Net SW IWV Cloud Albedo Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Radiation Attribution Attribution of Shortwave bias to surf albedo, clouds and IWV Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Radiation Attribution Radiation Attribution: Clouds dominate Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Cloud Regime Definition Which cloud regimes are responsible for this bias?? Cloud regimes, based on cloud occurrence at three levels of the atmosphere (Obs from ARSCL-Value Added Product)

CAUSES: Cloud Regime Analysis Composite cloud-related radiation bias into cloud regimes Width of the bars: Frequency of occurrence of regime Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Cloud Regime Analysis Composite cloud-related radiation bias into cloud regimes Width of the bars: Frequency of occurrence of regime Height of the bars: Mean cloud radiative effect Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Cloud Regime Analysis Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Cloud Regime Analysis Deep cloud regime problematic in most models, due to: Too small SW cloud radiative effect or Too small frequency

CAUSES: Cloud Regime Analysis Deep cloud regime problematic in most models, due to: Too small SW cloud radiative effect or Too small frequency Suppressed convective activity during daytime?

CAUSES: Surface Rain Analysis Deep cloud regime problematic in most models. Suppressed convective activity during daytime? Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Surface Rain Analysis Deep cloud regime problematic in most models. Suppressed convective activity during daytime? Most models are actually too wet in the afternoon! Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Surface Rain Analysis Deep cloud regime problematic in most models. Suppressed convective activity during daytime? Van Weverberg et al. 2017 JGR (submitted)

CAUSES: Relation to Warm Bias? Energy balance, given too much SW: Ma et al. 2017 JGR (submitted)

CAUSES: Relation to Warm Bias? Warmest models energy balance: too dry evaporative fraction too small Ma et al. 2017 JGR (submitted)

CAUSES: Relation to Warm Bias? Coolest models energy balance: too wet evaporative fraction too large Ma et al. 2017 JGR (submitted)

Recommendations to developers All models suffer from excess shortwave at the surface Mainly because of too infrequent or too transparent deep cloud Daytime precipitation rates/frequency actually overestimated, so not an issue of insufficient convective triggering! Possibly need to tune down precipitation efficiencies in midlatitudes to retain more cloud aloft and reduce rain rates Needs to be concerted with better captured elevated nocturnal convection, otherwise too dry overall, leading to too small evaporative fraction and maintaining the warm bias

CAUSES 4 CAUSES papers submitted: 1) Morcrette et al. (Near-surface temperature errors in NWP and climate model 5- day hindcasts near the Southern Great Plains) 2) Van Weverberg et al. (Attribution of surface radiation biases in NWP and climate models near the U.S. Southern Great Plains) 3) Ma et al. (On the role of surface energy budget errors to the warm surface air temperature error over the central U.S.) 4) Zhang et al. (Diagnosis of the Summertime warm bias in CMIP5 climate models at the ARM Southern Great Plains site) More info and data: http://portal.nersc.gov/project/capt/causes/